Neutrophil Resolvin E1 Receptor Expression and Function in Type 2 Diabetes

Over the past decades, the global prevalence of type 2 diabetes (T2D) has increased drastically. According to the International Diabetes Federation, roughly 415 million adults have diabetes, and the incidence of T2D is estimated to increase by 2040 (1, 2). This represents a global problem, heavily impacting quality of life, lifespan, and overall healthcare costs. In the natural course of T2D, it typically takes 15–20 y for hyperglycemia to progress to the establishment of disease. Uncontrolled inflammation plays an essential role in the pathogenesis of diabetes and its associated pathologies. Over the long-term, uncontrolled T2D can cause a cluster of diseases that are linked through inflammatory pathways, such as obesity, cardiovascular diseases, blindness, chronic kidney diseases, and periodontal diseases (3, 4).

Inflammation can play a protective role against injury and infections, but prolonged or excessive inflammation can lead to pathology (5). The resolution phase of inflammation is activated temporally, after an acute challenge, and involves eicosanoid class switching from proinflammatory to proresolution lipid mediators (LMs). The failure of inflammation to resolve leads to chronic oxidative stress, tissue damage, scar formation, and fibrosis (6–10). Continuous unresolved inflammation establishes a pathological response, leading to chronic disease (11).

A central characteristic of functional acute inflammation is a rapid return to homeostasis. A protective acute inflammatory response is tightly regulated by a genus of specialized proresolving LMs (SPMs). SPMs are produced via the actions of specific lipoxygenases on different substrates, including arachidonic acid (AA)–derived lipoxins (lipoxin A4 [LXA₄] and lipoxin B4 [LXB₄]), eicosapentaenoic acid (EPA)–derived E-series resolvins (RvE) 1–3, docosahexaenoic acid (DHA)–derived D-series resolvins (RvD) 1–6, maresins, and protectins (10, 12–24). Activation of resolution signals occurs when specific SPMs interact with cognate G protein–coupled receptors present on the cell surface (25, 26). LXA₄ binds its receptor, ALX/FPR2, transducing signals with potent actions in experimentally induced animal disease models (27). RvE1 (5S, 18R-trihydroxy-6E, 8Z, 11Z, 14Z, 16E EPA) binds to at least two G protein–coupled receptors, BLT-1 (a leukotriene B4 [LTB₄] receptor) and ERV-1 (formerly ChemR23), limiting neutrophil migration and accumulation and stimulating nonphlogistic recruitment of monocyte/macrophages for phagocytosis of apoptotic neutrophils and bacteria, which are eventually cleared through the lymphatics (16, 17). RvE1 binds to ERV-1 on monocyte/macrophages with high affinity (K_d = 11.3 ± 5.4 nM) and binds with lower affinity to BLT-1 on neutrophils (K_d = 48.3 nM) (1, 2). RvE1 treatment positively influences the outcome of inflammatory disease, such as murine colitis, periodontitis, and T2D, in animal models (28–37). Many studies focused on these LM ligands and their downstream functions, but there remains a dearth of research addressing how the receptors of resolution behave in chronic human diseases.

In T2D, uncontrolled inflammation has detrimental effects, and dysregulation of resolution is a possible link to the severity of disease presentation and therapy (38). In our previous work using a monogenic murine model of obesity and T2D, db/db mice exhibited decreased neutrophil chemotaxis, delayed wound healing, deficient phagocytosis, delayed neutrophil apoptosis, and defective clearance of inflammatory lesions (39). Deficient phagocytosis of pathogens promotes chronic diseases; for example, phagocytosis of Porphyromonas gingivalis, a key pathogen in periodontal disease (40), is deficient in diabetes (41). It remains unknown why diabetic immune cells fail to resolve inflammation. Because RvE1 increases cell functions in db/db mice, we investigated its cognate receptors, ERV-1 and BLT-1, on human neutrophils from volunteers with and without diabetes.

The goal of this study was to investigate the basis for inflammatory dysregulation in T2D. We found that the expression and function of a key receptor in resolution, ERV-1, was upregulated in T2D neutrophils. Signaling via ERV-1, not BLT-1, was predominant for RvE1 actions. TNF-α– and LPS-induced ERV-1 expression was modulated by RvE1. Deficient phagocytosis of diabetic neutrophils was rescued by higher doses of RvE1 compared with healthy neutrophils.

Subjects were recruited to the Center for Clinical and Translational Research at the Forsyth Institute, and samples were obtained under consent approved by the Forsyth Institute Review Board (protocol #11-03). Peripheral venous blood (∼60 ml) was collected from patients diagnosed with T2D and from healthy nondiabetic controls (39). The diagnosis of T2D was made by the subject’s primary care physician following American Association of Diabetes guidelines (42). HbA1c was used to determine 3-mo historic glycemic levels (Supplemental Table I). All blood donors were patients with either uncontrolled diabetes or healthy controls. Subjects that took insulin sensitizers, nonsteroidal anti-inflammatory drugs, or antimicrobials within 3 mo of the start of the experiment were excluded.

Human neutrophils were isolated from human whole blood by Ficoll-Histopaque density-gradient centrifugation (Histopaque-1077 and Histopaque-1119; Sigma-Aldrich). Neutrophils were isolated after isotonic lysis of RBCs, followed by two washes in PBS (Sigma-Aldrich). In culture experiments, neutrophils were obtained and incubated with RPMI 1640 medium (Sigma-Aldrich) supplemented with 10% FBS (v/v) (Life Technologies) at 37°C.

To confirm that isolated cells were neutrophils, cells were stained with Wright-Giemsa. After centrifugation at 2000 × g for 5 min, cell pellets were suspended in PBS (200 μl), cells were counted with a hemocytometer, and 50 μl of each cell suspension was mixed with 150 μl of 30% BSA in PBS, centrifuged onto microscope slides at 500 rpm for 5 min using a cytospin centrifuge, air-dried, and stained with Wright-Giemsa to identify individual cell type.

After cell isolation, neutrophils from subjects with T2D and normal controls were incubated with stimulants, including LPS (10 ng/ml), RvE1 (0.1–100 nM), TNF-α (10 ng/ml), and LTB₄ (10 nM). Various combinations of the agonists were explored. When two compounds were used, the first was incubated for 15 min at 37°C in 5% CO₂ before addition of the second.

To understand whether ERV-1 is functional in T2D, peripheral blood neutrophils were isolated and treated with RvE1 (10 nM) in the presence and absence of the BLT-1 receptor antagonist, U230495 (1 ng/ml), and an RvE1 receptor antagonist Ab (1 ng/ml). Various combinations of the RvE1 and antagonists were explored. When two compounds were used, the first was incubated for 15 min at 37°C in 5% CO₂ before addition of the second. The total stimulation time was 30 min.

To determine phosphorylation of ribosomal S6 (rS6), cells were plated into 96-well plates (100,000 cells per well). RvE1 (1 nM) was incubated with neutrophils for 30 min at 37°C, followed by cellular permeabilization (BD Cytofix/Cytoperm solution kit; BD Biosciences). Cells were labeled with allophycocyanin-conjugated anti–phospho-rS6 Ab (BD Biosciences) for 30 min at 37°C. Quantification of phosphorylation was determined by gating ERV-1⁺ cells only using a FACSAria II (BD Biosciences).

Recombinant human ERV-1 was transfected into Chinese hamster ovarian cells (CHO^ERV-1+; GenScript). ERV-1 receptor CHO^ERV-1− cells were obtained from Sigma-Aldrich. Cells were cultured in Ham’s F-12 medium supplemented with 10% FBS (Life Technologies), 100 U/ml Zeocin, and Hydromycin antimicrobials and maintained at 37°C in 5% CO₂. Cells were incubated with RvE1 for 1 h before total RNA was extracted.

RvE1 was prepared by total organic synthesis (16, 43). The structural integrity of RvE1 was monitored using UV tandem liquid chromatography–tandem mass spectrometry (LC-MS/MS). Immediately before use, RvE1 was diluted in PBS to a final ethanol concentration < 1%.

Peripheral blood neutrophils were isolated from healthy adults and those with T2D. Isolated cells were incubated with anti-Fc receptor (BD) blocking Ab (5 μg/ml × 10⁶ cells, 15 min) and then labeled with anti-human ERV-1 Alexa Fluor 488–conjugated Ab (10 μg/ml × 10⁶ cells, 1 h at room temperature [RT]) or anti-IgG Alexa Fluor 488 (isotype control; R&D Systems). Expression of ERV-1 on neutrophils was evaluated by immunofluorescence and quantified by flow cytometry. Cells were also stained with PE-conjugated anti-human CD11b, FITC-conjugated anti-human CD14, and allophycocyanin-conjugated CD18 Abs (10 μg/ml × 10⁶ cells, 1 h at RT) (BD Biosciences). Expression levels of the proteins were monitored by flow cytometry (FACSAria II; BD Biosciences) and analyzed with FlowJo software (Tree Star).

Total RNA was isolated from human neutrophils with TRIzol Reagent (Life Technologies, Carlsbad, CA) and purity was confirmed using a NanoDrop 1000 Spectrophotometer (Thermo Scientific). RNA was stored at −80°C in RNAlater. RNA was reverse transcribed using a High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems). mRNA expression levels were quantified by real-time PCR using SYBR Select Master Mix (Applied Biosystems) on a LightCycler 480 (Roche Diagnostics). Reactions were performed under the following conditions: preheating for 10 min at 95°C, followed by 45 cycles of denaturation for 5 s at 95°C, annealing for 10 s at 60°C, and extension for 6 s at 72°C. Relative gene expression was normalized to GAPDH. Data are expressed as Δ cycle threshold of mRNA levels (Fig. 1G); primers related to resolution and inflammatory genes are listed in Supplemental Table II. The following human ERV-1 primer sequences were used for CHO experiments: forward 5′-ATAGAATGGAGGATGAAGATTACAACACT-3′, reverse 5′-TCCCGAGGAAGCAGACGATG-3′. Table data are expressed as fold change.

Blood samples were collected and centrifuged at 2300 rpm, and serum was isolated and frozen at −80°C until analysis. Methanol (four volumes, 4°C, 30 min) containing 500 pg of deuterated internal standards d₄-LTB₄, d₈–5S-HETE, d₅-LXA₄, and d₄-PGE₂, to facilitate LM identification and quantification, was added to the serum. LMs were extracted using C18-silica reverse-phase cartridges and Biotage RapidTrace (44). Mediators were eluted using 6 ml of methyl formate, dried using TurboVap LV, and suspended in methanol/water for LC-MS/MS. The LC–UV coupled with MS/MS system includes a QTRAP 6500 equipped with a Shimadzu SIL-20AC autoinjector and an LC-20AD binary pump. An Agilent Eclipse Plus C18 column (100 mm × 4.6 mm × 1.8 μm) was used with a methanol/water/acetic acid gradient of 60:40:0.01 to 100:0:0.01 (v/v/v) at a flow rate of 0.5 ml/min. To monitor and quantify the levels of the specific LMs, a multiple reaction–monitoring (MRM) method with signature ion fragments for each molecule was used. Identification was conducted using published criteria (44). Calibration curves were obtained using synthetic and authentic LM mixtures, including d₈–5S-HETE, d₄-LTB₄, d₄-PGE₂, LXA₄, LXB₄, LTB₄, PGE₂, PGD₂, PGF_2α, thromboxane B2 (TxB₂), RvE1, RvE2, RvD1, RvD2, RvD3, RvD5, PD1, and maresin 1, at 12.5, 25, 50, and 100 pg. Linear calibration curves for each were obtained with r² values of +0.98 to +0.99. Quantification was based on peak area of the MRM transition and the linear calibration curve for each compound (44).

Human neutrophils were cultured in RPMI 1640 medium (Life Technologies) supplemented with 10% heat-inactivated FBS, 5.5 mM glucose, and 1% DMSO (Sigma-Aldrich) and incubated at 37°C. P. gingivalis strain A7436 was cultured on 2% Trypticase soy agar supplemented with 2.6% Brain Heart Infusion Agar (BD BBL), 1% (w/v) yeast extract (BD Bacto), 5% defibrinated sheep RBCs (Northeast Laboratory Services), 5 μg/ml hemin, and 0.5 μg/ml vitamin K (Sigma-Aldrich). All cultures were placed in an anaerobic chamber (85% N₂, 10% CO₂, 5% H₂ at 37°C). Colonies were transferred from the plate to Wilkin’s broth (Oxoid) and grown for 4 d. Bacterial titers were determined at 600 nm using a spectrometer (SmartSpec 3000; Bio-Rad) and adjusted to OD = 1.0 (∼130 CFU/ml) prior to experiments. Human neutrophils were counted and aliquoted (1 × 10⁶ cells). Bacteria were labeled with BacLight Green (Molecular Probes) for 20 min at RT with gentle agitation and washed twice with PBS. Labeled bacteria were opsonized in heat-inactivated normal serum (Sigma-Aldrich) for 30 min at RT and incubated with human neutrophils (1:20 ratio) for 1 h in serum and antibiotic-free medium at 37°C. Cells were gently washed, extracellular fluorescence was quenched by trypan blue, and phagocytosis was determined by flow cytometry using a FACSAria II (BD Biosciences). A similar protocol was followed for labeled zymosan fluorescent bioparticles (Life Sciences). Data are expressed as the phagocytic index (percentage of phagocytic neutrophils × mean fluorescence intensity [MFI]).

Results are expressed as mean ± SEM. Statistical analysis was performed using Prism 6 software (GraphPad). The Wilcoxon test and the Student unpaired t test were used to compare measurements. The p values ≤ 0.05 were considered statistically significant.

To investigate the role of the RvE1/ERV-1 axis in T2D, peripheral whole blood was collected from subjects enrolled in the study (healthy subjects, n = 83; T2D subjects, n = 83) after obtaining informed consent. The experimental protocol for this study was approved by the Institutional Review Board of the Forsyth Institute (IRB #13-07). Elevated serum glucose was evident in patients with T2D compared with controls (p < 0.0001, Fig. 1A), and HbA1c levels confirmed serum measurements (Supplemental Fig. 1). Serum cholesterol, body mass index (BMI), and age were higher in subjects with T2D (p < 0.05, Fig. 1B; p < 0.0001, Fig. 1D; p < 0.0001, Fig. 1E, respectively). Total neutrophil counts were increased in subjects with T2D (p < 0.05, Fig. 1C), but no significant difference was found in total monocyte counts (p = 0.176, Fig. 1F). Healthy volunteers matched for gender and race served as controls. Demographic parameters are reported in Supplemental Table I. Correlation analysis showed a positive association between age and cholesterol (p = 0.006). Negative associations were found when evaluating ERV-1 and BLT-1 receptors, BMI, and glucose levels. Subgroup analysis of BMI and age by Pearson coefficient showed no significant association between BMI and age (p > 0.05). The gene-expression profile of neutrophil mRNA was evaluated. Resolution and inflammatory genes were plotted as a heat map of individual samples (Fig. 1G). The primers and respective genes investigated in this study are listed in Supplemental Table II.

To elucidate the expression profile of ERV-1, mRNA and protein levels were evaluated. mRNA expression levels of ERV-1 were quantified by PCR (Fig. 1G) and protein expression levels of ERV-1 on neutrophils from individuals with T2D cells were isolated and quantified by flow cytometry (Fig. 2A). The ERV-1 receptor was only known to be active on monocytes/macrophages and the BLT-1 receptor was only known to be active on neutrophils (29). To determine the cell surface expression patterns of the ERV-1 receptor on human neutrophils in T2D, isolated cells were labeled with anti–ERV-1 Ab for quantification by flow cytometry and immunofluorescence (Fig. 2A). ERV-1 expression was increased significantly on neutrophils isolated from patients with T2D (Fig. 2A, p < 0.0001; immunofluorescence Fig. 2D). The distinct coexpression pattern of neutrophil ERV-1 and BLT-1 receptors demonstrates that there is high expression of ERV-1 and low expression of BLT-1 in T2D (Fig. 2B, p < 0.001), whereas there is a lower expression of ERV-1 and a higher expression of BLT-1 in healthy individuals (Fig. 2A, 2B). CD11b, CD14, and CD18 expression by neutrophils from T2D and healthy controls is not statistically different (Fig. 2C–F), demonstrating a specific regulation of RvE1 receptors. Double-positive (ERV-1⁺/BLT-1⁺) neutrophils were significantly increased in T2D, whereas double-negative neutrophils were not (Fig. 2G, 2H). Double-positive or double-negative CD11b/CD14 neutrophils were not different between groups (Fig. 2I, 2J).

Stimulation of human macrophages transduces ERV-1 signals through the Akt/rS6/mTOR pathway (36). In this study, we demonstrate that neutrophils are also stimulated with RvE1, ERV-1, and BLT-1 signal transduction through an identical pathway. We monitored phosphorylation of rS6 protein after alternatively blocking each receptor to determine which receptor(s) is active in T2D neutrophils. We found that, when ERV-1⁺ neutrophils are incubated with RvE1 (10 nM), intracellular rS6 signaling increases compared with baseline controls (Fig. 3A). Because of the overexpression of the ERV-1 receptor and rS6 phosphorylation in T2D, we investigated the baseline rS6 phosphorylation levels of unstimulated cells. We quantified levels of phospho-rS6 in neutrophils from T2D and healthy controls (p < 0.05, Fig. 3B). RvE1 treatment significantly increased rS6 phosphorylation in T2D and healthy neutrophils (p < 0.001, Fig. 3B); expression in T2D was significantly higher (p < 0.001, Fig. 3B, 3D). Neutrophils were stimulated with TNF-α or LPS and blocking agents (ERV-1–blocking Ab and the BLT-1 antagonist U230495) to assess the functional transduction of signals (Fig. 3C). Our findings indicate that RvE1 activates rS6 phosphorylation through BLT-1 in neutrophils from healthy donors and through ERV-1 in neutrophils from T2D subjects. Treating neutrophils with both blockers simultaneously completely eliminated rS6 phosphorylation in neutrophils from T2D and healthy subjects, suggesting that BLT-1 and ERV-1 are the key receptors that transduce RvE1 signals.

RvE1-stimulated healthy neutrophils were shown to signal via BLT-1 and inhibit LTB₄ function through competitive inhibition (29). Because surface BLT-1 is downregulated and ERV-1 is upregulated in T2D, we performed experiments to determine how ERV-1 expression is regulated. We exposed T2D and control neutrophils to the proinflammatory cytokine TNF-α or bacterial LPS and then measured surface expression of ERV-1 by FACS. When neutrophils were treated with TNF-α (10 ng/ml) for 1 h, ERV-1 protein expression increased in T2D and healthy controls, with significantly greater expression in T2D (p < 0.01, Fig. 4A).

Interestingly, LPS stimulation did not upregulate ERV-1 on healthy neutrophils (a slight reduction was observed), but it significantly upregulated ERV-1 on T2D neutrophils (p < 0.01, Fig. 4B). Neutrophils were also exposed to RvE1 15 min before or 15 min after TNF-α or LPS. In each case, RvE1 treatment of neutrophils returned ERV-1 surface expression to the levels of unstimulated healthy neutrophils (Fig. 4A, 4B). When cells were treated with LTB₄, patterns remained similar to baseline, with a slight increase in healthy neutrophils. A 15-min pretreatment or posttreatment with RvE1 decreased ERV-1 expression (Supplemental Fig. 2). Because the kinetics of ERV-1 clearance and re-expression are known (16, 29, 36, 45, 46) and are not likely to account for downregulation of the receptor, we investigated transcriptional regulation to determine whether expression was controlled at the RNA level. We transfected CHO cells with the human ERV-1 receptor (designated CHO^ERV-1+). Expression of ERV-1 protein was confirmed compared with CHO^ERV-1− by immunofluorescence (Fig. 4C) and by quantitative PCR (Fig. 4D). We further confirmed that regulation was a direct response to RvE1 in a dose-response experiment (1–100 nM RvE1). Lower concentrations of RvE1 (1 nM) reduced the overexpression of ERV-1 on CHO^ERV-1+ cells (nonsignificant, Fig. 4E), whereas 10 and 100 nM RvE1 significantly reduced ERV-1 expression (both p < 0.01, Fig. 4E).

Sera were subjected to LC-MS/MS profiling to understand the profile of LMs and SPMs in peripheral blood from T2D and healthy subjects. Selected bioactive metabolomes of DHA, EPA, and AA were analyzed (Fig. 5A, 5B). MS/MS fragmentation spectra were used for identification (Fig. 5C). LMs from the DHA bioactive metabolome (RvD1–6, aspirin-triggered RvD1, protectin D1, and maresin 1) and the EPA bioactive metabolome (RvE1–3) were identified and quantified following established criteria (47). In these sera, we also identified and quantified mediators from the AA bioactive metabolome, including LXA₄, LXB₄, LTB₄, and its further metabolites 20-OH–LTB₄, 20-COOH–LTB₄, PGD₂, PGE_2, and PGF_2α, and TxB₂. Quantification of mediators from the EPA metabolome in healthy sera revealed RvE1 at 0.6 ± 0.001 pg/ml; RvE2 and RvE3 were below the limits of detection (Fig. 5). In T2D sera, a slight decrease in RvE1 levels was observed (0.4 ± 0.1 pg/ml), and RvE2 and RvE3 were similarly below the limits of detection. Because endogenous RvE1 is a key target of this study, we also measured it by ELISA (Supplemental Fig. 1). Results demonstrated a decrease in the LM in T2D (p < 0.001, Fig. 5D). The AA bioactive metabolome caused a marked increase in LTB₄ in serum from T2D (16.8 ± 4.0 pg/ml) compared with healthy controls (5.8 ± 2.1 pg/ml, p < 0.05, Fig. 5B). TxB₂ was higher in healthy serum compared with T2D serum (148.5 ± 63.9 and 63.9 ± 37.7 pg/ml, respectively, p < 0.001, Fig. 5D). An increase in inflammatory mediators and a slight decrease in resolution mediators were found in the sera from T2D subjects compared with control subjects (Fig. 5).

SPMs, like RvE1, have the ability to influence cell behavior and act as biochemical agonists, altering functional responses in infection and inflammation (48–52). Phagocytosis is a key biological process for neutrophils and macrophages in the initial host response, promoting clearance and return to homeostasis. We reported a deficiency in the phagocytosis of the pathogen P. gingivalis in T2D (39, 41). To understand the functional actions of neutrophils in T2D, we incubated cells with RvE1 and assayed phagocytosis in vitro with bacterial pathogen bioparticles and zymosan fluorescent bioparticles. In a kinetic assay, T2D neutrophils exhibited decreased phagocytosis after 1–4 h (Fig. 6A, upper panel). Neutrophils were pretreated for 15 min with 1, 10, or 100 nM RvE1, and activation of ERV-1 and phagocytosis were assessed (Fig. 6A, lower panel). Compared with controls, diabetic neutrophils responded differently to RvE1 pretreatment in a dose-dependent manner (Fig. 6A). The results were consistent when zymosan bioparticles were used (Fig. 6B–D). For neutrophils from healthy subjects, low-dose (1, 10 nM) RvE1 enhanced phagocytosis over baseline in a dose-response manner. Consistent with signaling results, phagocytosis is mediated through ERV-1 receptor activation in diabetes and through BLT-1 in healthy neutrophils (Fig. 6D). The 100-nM dose of RvE1 had no additive impact on normal neutrophil phagocytosis, suggesting an optimal dose range. For diabetic subjects, higher doses were required to achieve a similar response (Fig. 6).

Neutrophils are essential cells for responding to inflammation under normal health and T2D conditions. Dysregulated immune cell functions are clearly seen in T2D, but the molecular basis remains ill-defined. We demonstrate that uncontrolled T2D impacts the phenotypic expression of the RvE1 receptor, ERV-1, on peripheral blood neutrophils. ERV-1 expression was originally reported as an active signaling receptor on macrophages and dendritic cells (53); however, the expression and function of ERV-1 on neutrophils in T2D and other chronic inflammatory diseases have not been investigated. Our findings show that ERV-1 expression is low on healthy human neutrophils, and BLT-1-mediated signaling is dominant. Neutrophils isolated from people with poorly controlled T2D express markedly elevated levels of ERV-1 on their surfaces that is functionally transducing signals, whereas expression and signaling through BLT-1 are significantly decreased (Figs. 1G, 2A). T2D neutrophils exhibit a markedly reduced capacity for phagocytosis of bioparticles and pathogenic bacteria that can be rescued with RvE1 treatment, but the concentrations of RvE1 necessary for activity far exceed those required for normal neutrophils from healthy subjects. Metabololipidomics of serum LMs revealed a marked increase in LTB₄ and a decrease in RvE1 as well as other SPMs (Fig. 5). Exogenous LTB₄ did not significantly change the baseline levels of ERV-1 on neutrophils in vitro.

Development of the metabolic syndrome prior to the onset of T2D is characterized by changes in the metabolism of glucose and fatty acids that activate innate immune responses that give rise to systemic insulin resistance and establish a state of chronic systemic inflammation (2, 4, 38). As a result, T2D is associated with deficiencies in the clearance of microbial infections, impaired phagocytosis, and chronic inflammation. In this study cohort, individuals with T2D exhibited marginally increased neutrophil counts, high cholesterol, and high glucose levels (Fig. 1). Increased inflammation in T2D is associated with prolonged activation of neutrophils, as demonstrated by the gene-expression profile (Fig. 1G), consistent with an inefficient clearance of bacteria that consequently increases susceptibility to infection. Additionally, inflammation was implicated in other complications of T2D, including cardiovascular disease, retinopathy, and a high incidence of oral manifestations, such as periodontal diseases and multiple abscesses (39).

The exact mechanisms by which neutrophils fail to activate resolution cascades in T2D remain unknown. Chronic inflammation in diabetes and/or obesity influences cell phenotype. In this study, subgroup analysis of BMI revealed no association with ERV-1 or BLT-1 receptor expression. We demonstrated previously that exogenous application of RvE1 in diabetic mice increased cellular functions, but the reason for the failure of resolution remained unclear. Neutrophils are central to the initiation and resolution phases of the acute inflammatory response, and both phases are tightly regulated by proresolution mediators; however, receptor expression and function in chronic diseases are yet to be explored. We reported previously that human neutrophils express ERV-1 (54); consistent with our findings (55), this was confirmed. In this study, we demonstrate that the RvE1 receptor ERV-1 is upregulated by neutrophils from people with poorly controlled T2D (Figs. 1G, 2A). This observation was specific to neutrophils, because monocytes did not show similar increases. The signaling pathways activated in neutrophils in response to RvE1 were inferred from work with monocytes/macrophages (36) but had not been demonstrated directly. We confirm that normal peripheral blood neutrophils also signal with rS6/mTOR pathways, primarily through the BLT-1 receptor. T2D neutrophils signal via rS6 as well; however, the main receptor is ERV-1. Interestingly, another ERV-1 ligand, chemerin 15, actively signals in chronic coronary lesions, but chemerin did not regulate the expression patterns of the receptor (49). There is a precedent for gain of function of ERV-1, depending on the macrophage phenotype. As reported recently, ERV-1 expression was found on functional M1, but not M2, macrophages (56). Thus, a proinflammatory phenotype favors high expression of ERV-1 on neutrophils. We also considered the possibility that the significant difference in age among our subjects (Fig. 1E) was a confounder in the ERV-1 receptor profile. In previous studies, aging was shown to influence resolution lipids, but no information was available on resolution receptors (47). Using Pearson’s correlation analysis, no positive association was found between age and ERV-1 and BLT-1 receptors, BMI, or serum glucose (p > 0.05). Age showed a positive correlation with cholesterol levels (p = 0.006).

The hypothesis that increased inflammation was the main driver of ERV-1 expression was tested. LPS and TNF-α activation of neutrophils resulted in increases in ERV-1 expression (Fig. 4A, 4B) similar to that seen at baseline for neutrophils from T2D; the increase was additive to the upregulation of neutrophils compared with healthy controls. This agrees with prior studies showing that neutrophils treated with TNF-α, the formyl-methionyl peptide FMLP, and IL-8 rapidly upregulated ERV-1 (55). Exogenous addition of RvE1 downregulates ERV-1 overexpression that was endogenously elevated in T2D or overexpression induced by LPS and TNF-α. We also found that neutrophil ERV-1, but not monocyte or lymphocyte ERV-1, was upregulated (55). Although previously published results demonstrated that anti-inflammatory ligands, such as chemerin, annexin A1, and α-melanocyte stimulating hormone, did not influence ERV-1 surface expression, RvE1 was able to regulate receptor expression in a dose-dependent manner (Fig. 3). The key question of whether ERV-1 expression was dependent on RvE1 acting directly on the receptor was addressed by expressing this receptor in CHO cells. When CHO^ERV-1+ cells were incubated with RvE1 (1–100 nM), mRNA expression decreased (Fig. 4E). This indicates the dose-dependent influence of RvE1 LM ligand on ERV-1 receptor levels.

The activation of ERV-1 signaling by RvE1 activates rS6 phosphorylation that turns on cellular biological functions, including increased phagocytosis and reduced NF-κB–induced cytokine production. Upon exogenous RvE1 activation at therapeutic doses considerably higher than those found in serum, neutrophils responded with rS6 phosphorylation (Fig. 3B). LPS and TNF-α treatment influenced ERV-1 expression and rS6 signaling (Fig. 3C). Upon exogenous RvE1 activation, overexpression of the receptor and increased phosphorylation were rescued. Understanding the molecular mechanisms that regulate the expression and function of the resolution receptors ERV-1 and BLT-1 is relevant to exogenous RvE1 biological functions.

It is plausible that RvE1 increases the rate of phagocytosis of healthy neutrophils; however, because ERV-1 expression levels were differentially regulated, the concentration needed to activate T2D neutrophils was unknown. The deficient phagocytosis observed in T2D is an obvious hallmark, and the formation of ulcers and periodontal abscesses is a major consequence of this deficiency. In this study, RvE1/ERV-1 binding activated phagocytosis in a dose-dependent manner, with higher responses when 100 nM RvE1 was added to T2D neutrophils. In healthy individuals, the effective dose (1–10 nM) is logarithmically lower than in T2D (Fig. 6A, 6B). Interestingly, at the highest dosage in healthy neutrophils, the effect of RvE1 was attenuated. At the highest dosage in T2D neutrophils, phagocytosis reached the peak response of normal neutrophils at a 10-fold lower dosage (Fig. 6).

These findings demonstrate that loss of regulation of inflammation is a biological event common to human chronic diseases, including T2D. Understanding deficient pathways in the resolution of inflammation is critical for designing new therapeutics for treating individuals with T2D. The traditional focus for managing T2D has been the control of hyperglycemia and insulin, not the resolution of inflammation. Receptors of inflammation resolution should be included as potential markers for cell phenotype and possible therapeutic targets. The results of this study demonstrate that ERV-1 is responsive to inflammatory stimuli and that the LM ligand RvE1 improves neutrophil-mediated inflammatory responses in uncontrolled T2D. Importantly, these results suggest that it is possible to correct deficient responses to RvE1 with added exogenous RvE1. Further studies are necessary to characterize ERV-1 receptor dysfunction in other models of inflammatory disease. Understanding these pathways and their regulation of disease is important for personalized medicine approaches for the treatment of inflammatory diseases.

We thank Paul Chung for excellent technical assistance and Laurie K. McCauley for helpful discussions. This work was performed at the Forsyth Institute and at the Lipid Mediator Metabololipidomics Core, Brigham and Women’s Hospital, Harvard Medical School.

The authors have no financial conflicts of interest.